Long-standing readers of Commsrisk will be conscious of how few articles articles now refer to revenue assurance. This might seem odd; I wrote a book on revenue assurance for comms providers in collaboration with other contributors to Commsrisk. However, it is difficult to keep writing new articles about a domain when nobody working in that domain has anything new to say about it. So I was momentarily pleased to see that KPMG India published the report of a new survey on revenue assurance and fraud management (RAFM). I hoped it might say something new about revenue assurance. Experienced risk managers can already calculate the likelihood that I was disappointed.
I have no desire to spend a lot of time discussing the flaws of this report because so little will be gained by the exercise. It might have been best to simply ignore it. This report changes nothing. There will still be some people working in telecoms revenue assurance next year just like there will still be some blacksmiths next year. Horses continue to exist; there remains a need to manipulate metal so their shoes fit properly. CDRs continue to exist; some will continue to go astray. However, nobody is expecting exciting revenue growth in the blacksmith sector, and nobody is anticipating major new breakthroughs in the methods used by blacksmiths. The same prospects apply to revenue assurance. Progress continues to be made in the science of materials; blacksmiths are not responsible for that progress. Much can and should be done to increase returns on investment and profit margins by using data to eliminate waste, improve efficiency and make better decisions; revenue assurance professionals are unlikely to be responsible for that progress.
By accident or design, revenue assurance was shunted into a silo where it has stagnated. Stagnation is good, if you do not want the burden of learning new skills and undertaking new challenges from one year to the next. Somebody who checks CDRs for 15 years in a row should not really be disappointed if their CDR-checking skills and experience has dropped in value because of changes to business models that they did not pay attention to. There is nothing wrong with being a blacksmith. However, blacksmiths are not going to improve their careers by reading reports that pretend there are exciting growth opportunities and recent breakthroughs in the methods they may use.
Normally I review reports by analyzing their contents. In this case, the review should discuss what was missing from the report because those omissions are more significant than any of the contents. Risks can only be rationally managed if we are conscious of the greatest risks. Revenue assurance stagnated because it remained so steadfastly focused on specifics that declined in importance relative to newer objectives and different risks that businesses faced. It concentrated on making horseshoes despite the rising number of cars. So here, in no particular order, are the most serious omissions from KPMG’s report.
There Is No Fraud Management in This RAFM Report
Despite being labeled as a ‘revenue assurance and fraud management’ report, there would have been no difference to its content if the authors stated that were only surveying revenue assurance activities. This is because they have ignored the principles of risk management. In risk management, we begin with a rational assessment of risk and then we examine potential mitigations. This report is written from the perspective of people who have already determined which methods they are trying to sell to customers, and who then spend their time justifying why those methods are needed. Put simply, the purpose of this report is to advertise what KPMG India wanted to advertise. That only involves methods to analyze the data which is already available to users. It says nothing about obtaining other intelligence which is not readily available, which is a key challenge in fraud management.
Nothing is said about obtaining qualitative judgements of risk when no quantitative data exists. This is an especially common challenge when dealing with new risks because there can be no expectation of relevant historic data. This report is full of premature references to GenAI and other fancy technologies like blockchain, but it says absolutely nothing about whether the business of running a comms provider has changed in the past, is currently undergoing change, or will change in the foreseeable future. There is no anticipation of where and how demand for risk management products and services arises. The authors are solely interested in the products and services they would like to supply. Revenue assurance is in the doldrums because it became synonymous with products that businesses did not want. KPMG has perpetuated the mistake of trying to change the customer to suit the product instead of changing the product to suit the customer.
Competent fraud management should include threat analysis; threats change over time and new threats emerge. The analysis of threats will be inadequate if it is only based upon data that is already available in a format that is suitable for automated scrutiny. That is why fraud management in the comms sector has not declined in the way revenue assurance has. The obvious need for intelligence about new threats creates the incentive to collaborate in the sharing of information and this then prompts the development and adoption of new mitigations that were not previously required. Fraudsters will innovate. They spur innovation with the community of comms providers and their suppliers.
Fraud managers are not perfect at gathering intelligence and sharing it with each other, but they can justify expenditure on collaborative and educational activities that would otherwise not occur. Or to put it another way, revenue assurance professionals from different comms providers have been meeting less and less frequently because they gain so little from speaking to each other. They are being propped up by events that claim to be for an RAFM audience but which have very little RA in them. This report is a different example of the RA community borrowing from fraud management — using the label to generate interest in the report — while refusing to learn the most important lesson. Nobody will pay for new RA tools and methods if nobody can show why they are needed.
An Inward-Looking Report That Ignores the Biggest Single Opportunity for RAFM Professionals
If I asked anybody anywhere — not just people who work for telcos — what they think of risks involving Indian telcos, then the most common answer will be “don’t know” and the next most common answer will be “scams”. That is what people who know nothing about telcos and know nothing about India think about when they think about Indian telecoms. Also, Indians are aware of scams and they hate them too. So it shows a desperate lack of imagination that a report from KPMG India that tries to increase RAFM sales not only pretends to cover fraud management without really covering it, but also fails to notice that the single biggest global driver of new investment in fraud management is the mitigation of consumer scams.
The report says literally nothing about the methods telcos are using to prevent consumer scams. The authors have completely ignored a profound opportunity to dramatically upscale the work of RAFM teams even though India is home to some of the most exciting new work in consumer fraud mitigation. The sad truth is that much of the mitigation of consumer scams will be orchestrated by other parts of comms providers because RAFM teams are not adjusting their mission to address the highest priorities of the business.
This Survey Is Not a Survey
The omissions from this report become easier to understand when you realize it is only pretending to relay the results of a survey. There are plenty of graphs. I do not trust any of them.
There is a section titled ‘Survey Methodology’ but it is just waffle. It reads like something ChatGPT would generate if you told it to bluff about survey methodology when there is no survey methodology. It does not say anything about the actual methods used to conduct the survey. At the very least, a trustworthy survey should tell the reader how many people were interviewed, what kinds of people were selected for the interviews, and what questions they were asked. None of that is stated in this report.
I expect the reason for this vagueness is that insufficient rigor was applied when choosing who would be interviewed, and insufficient discipline governed the interactions between the surveyors and the interviewees. It is not possible to describe a methodology if there is no consistent method. And the likely explanation for a lack of consistent method is that the people interviewed had too little in common to reliably answer the questions that were put to them.
Some people doing work in RA have a job title which refers to RA; others do not. Some people work in joint RAFM departments; other businesses split RA and fraud management. Some people reduce fraud by interviewing employees of their own business and firing the ones who committed fraud; others examine CDRs. It is quick and easy to assemble groups of ‘RAFM’ professionals where one person in the first telco has something vaguely in common with a second person from a second telco, who has something vaguely in common with a third person from a third telco, who has something vaguely in common with a fourth person from a fourth telco, but where the first and the fourth person have nothing in common. If you try asking them the same questions then the results are garbage. And if all of the most important questions are really about increasing sales of AI, then you are guaranteed to get garbage because very few of these people will really need AI, and even fewer are currently using AI.
Sometimes We Should Ask People about People, and Not Just Ask Them about Machines
You already know the purpose of KPMG India’s report without seeing it: they want to promote the sale of technology. There are genuine reasons to use technology to mitigate risks, optimize revenues, and all those other good things. But in my experience, there has never been a shortage of people arguing that businesses should spend more on technology. Businesses underinvest in RAFM technology because they underinvested in RAFM people. If somebody does not understand technology they generally will not buy it, and if they still buy it then they will make a mess of using it.
Selling technology without training people how to do their job is like selling cars without teaching people how to drive. Some will muddle through. Others will cause a lot of pain. And if you have never learned how to drive a car then it would be an even bigger mistake to hand you the keys to a 12-wheel truck or ask you to pilot the maiden flight of an Airbus A380. Having bigger tech, better tech, more expensive tech is no guarantee you will get to your destination sooner if there is nobody competent to steer the machine in the right direction. This report conspicuously ignores the single most important cause of the stagnation of revenue assurance: the failure to invest in the training and development of revenue assurance professionals.
Conclusions
This report was essentially an advert for tech. I thought revenue assurance had already stagnated to the point where everybody had stopped pretending that all the problems of RA will be solved by spending more on tech. Some spending on training people to do revenue assurance would be a good idea. A route for career progression would help with retaining and motivating good staff. But this was an advert for tech because… AI, AI, AI!?!! I guess the authors do not care if a blacksmith will be put in the cockpit so long as they can talk about the advanced features of an Airbus A380.
Anybody attending a conference about telecoms fraud in 2025 will have been bombarded with hype about the use of AI to tackle consumer scams. However, the authors of this report lacked the imagination to link their desire for AI with the single most pressing driver of new anti-fraud sales. They appear to have literally copied the same business case that RA vendors were using 10 years ago except that they added many references to AI without explaining why AI might be beneficial or which of those 10-year old problems could most effectively be solved with AI. They might as well have written a report claiming blacksmiths should use AI to make horseshoes.
One good thing about AI is that it can learn. This gives AI an advantage over people who refuse to learn. This woeful report from KPMG India suggests that nothing has been learned from the last 10 years of RA stagnation. It has the same excessive bias towards new technology as we saw then. How many telco employees have truly benefited from that bias in the meantime? The obsession with buying systems left the employees stagnating. They were still expected to hammer away at errors all day long, although sometimes they were given a different anvil. The budget for technology would sometimes shoot up; the wages for the people using it never did.
If there is no reward for the employee’s imagination, for the employee’s creativity, for insights that only a human employee can bring to their work, then we end up with employees who have no imagination, no creativity, and no insights to share. We end up with staff who actually believe the dreck in a report like this because they have never been encouraged to engage in critical thinking. Somebody needs a spark of insight if they are to persuasively argue for spending on those few novel technologies that would actually benefit the business. A lack of imagination leaves only one convincing business case for new technology: the cutting of costs. Buy technology to replace people. And what happens when the technology has replaced the last employee? Who will be reading these reports then?
I do not recommend that you read it, but KPMG India’s report on RAFM, dated August 2025, is freely available here.
